# Appearing voltage when connecting different grounds

We have designed instrumentation equipment. When several instrumentation units are connected their grounds must be connected. The first proposal was to connect the GNDs by cascade topology. But we have the problem of connecting different grounds: each one has different voltage instead the same reference. So, equipment behaviour becomes imprecise. It doesn' t works correctly. I know this problem is a typical one. This missfunction has been corrected by using the common gnd source supply pin for all the boards gnd instead being cascade wired. But I would like to improve the pcb design in order to save the previous idea.

I'm trying to remember the school subject in which one how to solve different ground voltage when its expected to be the same. There was some circuit design strategies to avoid the problem. But It stills so far from my memory.

Is it possible to connect gnd pins having the same reference? If you start to design a circuit that is expected to be connected with other by connecting gnd, how do you face this. What kind of pcb and circtuit design strategies will you take into account?

It will be grateful if some one could give me a little strategy list or source where to find the current strategies.

In ex, here:

https://electronics.stackexchange.com/a/39477/144534

Regarding "A well designed circuit must have a reliable ground, which means that the 0 V at one point should be as close as possible to that 0 V at any other point of the ground net. Zero difference is not always possible if you're working with high currents, but the difference should be as low as possible."

Some suggestions about how to get this could be helpful.

• Very hard to answer without knowing nature of the circuits. Are differential signals between the two an option? Nov 16, 2018 at 8:35
• What currents do you have flowing in the GROUND? 1 milliamp DC? 100 amps at 50Hertz? currents from a 10 horsepower motor, with the commutator sparks making the problem very difficult? MicroController 32-bit-data-buss return currents with 1nanoSecond edges? Nov 16, 2018 at 10:59
• I will take your comments into account. I hope I could give you the answer soon. Nov 20, 2018 at 13:24

## 2 Answers

If your system draws current from a source with a cable with impedance R_cable, you may have a difference between the ground of two system, the maximum voltage is R_cable * I_peak

This problem is related to wiring but you can compensate it in your design.

Methods can improve your design

• use differential data transfer instead of transferring analog or TTL/CMOS voltage

• Isolate your power system, and install ground between isolated circuit

simulate this circuit – Schematic created using CircuitLab

when the isolated circuit gets a common GND2 signal, they don't transfer power current from this wire and your design is stable for signaling.

• transfer current instead of voltage, for example, 4_20mA (current can compensate in the transmitter)

In some cases, you can compensate for the error but the circuit and cable length should be fixed.

There are many methods you can use, some of them presented here.

Cascading ground is not an option, never. You collect all GNDs and bind them together into a single point. Then you connect this single point to a reference- earth potential. It has to noted that not a single GND is already connected to the earth potential. If such device exists, then it becomes the reference.